Genetic and phenotypic heterogeneity in autism and autism spectrum disorders (ASD) pose significant challenges for research focused on defining biomarkers, developmental trajectory and treatment and outcomes. This provides a key rationale for requiring Fragile X testing and high resolution karyotyping using SNP arrays on all subjects enrolled in ACE Centers or Networks, as we will continue to in this ACE. In our previous ACE Center Project II, we hypothesized that studying ASD endophenotypes, would aid in identification of more homogeneous patient subgroups and hasten identification of genetic loci. We showed how a common ASD susceptibility variant in CNTNAP2 modulates brain function, connecting gene to brain to endophenotype for the first time in ASD. We also identified several cases of rare, large copy number variation (CNV) and smaller variants of less certain pathogenecity. Here we propose to continue to genetically characterize all ACE probands, hypothesizing that identifying certain etiological subclasses may provide more homogeneous populations that will be more predictive of trajectory and outcome. We will integrate identification of CNV with gene expression data to identify dysregulated genes within and near CNVs, thus improving classification of pathogenecity, and identify those with mutations currently undetectable by structural variant analysis alone. We will take a systems approach to functionally group these genes into biological pathways, and thus to group patients by shared molecular defects. We will then relate shared molecular pathway defects in the patient subsets to the phenotypic biomarker measurements collected in projects l-IV. We will test the relationship between known and newly discovered genetic variants and measures of behavior, eye-tracking/pupillometry, EEG, and brain imaging at both single time points and examining longitudinal trajectories, as well as their influence on response to treatment. In this way we seek to connect genetic variation to measures of brain function as a means of unraveling the genetic and phenotypic heterogeneity observed in ASD, and to develop improved predictors of diagnosis and treatment response.